1. Field of the Invention
The present invention relates to an image processing apparatus, image processing method and image processing program, and more particularly, to an image processing apparatus, image processing method and image processing program that synthesize into a document image file character, graphic and photograph areas that have been respectively extracted from image data obtained via reading of an original document and that have been subjected to image processing.
2. Description of Related Art
Image data obtained via reading of an original document using a scanner or other device is large in size and is not appropriate for storage or for transmission and receipt as is. The image data therefore is subjected to appropriate image processing, such as compression, depending on the type of the image. However, because image data that is obtained via reading of an original document includes character areas comprising character images, graphic areas comprising graphic images and photograph areas comprising photograph images, the problem exists that when the image data is subjected to irreversible compression appropriate for the photograph areas, although the data size is reduced, the characters become difficult to read, while if the image data is subjected to compression appropriate for the character areas, the compression ratio becomes small. Accordingly, an image processing apparatus is known that, after separately extracting character, graphic and photograph areas from image data obtained from an original document and performing compression appropriate for each type of area, creates a document image file by synthesizing these post-processing areas. Using this image processing apparatus, the size of the image file can be reduced while maintaining image quality.
However, the problem then arises that because the extraction of the various areas by the image processing apparatus described above is generally performed based on circumscribed rectangles for each area in order to avoid complex processing, some of the extracted areas partially or completely overlap, depending on the locations of the areas, resulting in missing image parts and thus in failure of the synthesized image file to reproduce the original.